PS2 Development with AthenaEnv

For the better part of a year, I have been pulled toward game development. About a decade ago I tinkered with Unity (and Blender) through a YouTube playlist, built one small game, and stopped there. A couple of years later, I got my hands on a Phat PAL PS2 and a FreeMcBoot memory card.

That sparked a question: what is it like to write game code that runs close to the metal, without the usual modern runtime layers? I first explored the PS2 SDK through the ps2dev/ps2dev Docker image and some YouTube tutorials. I was writing C++ again after a long break, but I did not get far. I was mostly feeling my way through the toolchain, so I shelved it after a few weeks.

Later, I found a video showing how to port a web game (JavaScript) to PS2 using AthenaEnv. That felt like the right path back into PS2 homebrew for me, especially with faster iteration than a full C++ workflow.

What is AthenaEnv?

AthenaEnv is a framework for building PS2 homebrew software with JavaScript. It provides an abstraction layer over core system interfaces, which makes early experimentation much faster than starting directly from low-level C/C++.

AthenaEnv uses a modified QuickJS interpreter. At a high level, it acts as a JavaScript loader. It runs main.js by default, but you can pass another file as the first argument when launching the ELF.

Enough theory. Time to build.

Setting up the project

To get started, I used the default template from this repository. For quick testing, I also installed the PCSX2 emulator so I could iterate without booting real hardware for every run.

First Test Case: 2048

As a test case, I wanted to evaluate how quickly I could get a browser game running on PS2. I used Claude to assist with an initial port of this JavaScript 2048 project (repo) to AthenaEnv.

Below is what the original game looks like in a browser:

The first generated version was rough, but after a few iterations I got to this state:

I still need to debug startup/input behavior. It might be a keybinding mismatch in the emulator. After that, I want to validate the same build on actual hardware before sharing an ISO.

What I learned

• A JavaScript-first workflow lowers the barrier for PS2 experimentation.
• Emulator testing is great for iteration, but hardware validation still matters.
• LLM-assisted ports can accelerate first drafts, but manual debugging is still the real work.

Next Steps

I can run ELF files on both console and emulator, but the end goal is packaging each project as an ISO. That means I need a clean Windows-first toolchain for image generation and testing.

I also need more reps on this platform before building a game from scratch. For now, I will keep porting small projects, documenting what breaks, and tightening the workflow.